Sign in with
Sign up | Sign in

Test Settings

System Builder Marathon, June 2010: System Value Compared
By

March 2010 System Builder Marathon Components 
 $550 PC$1,000 PC$2,000 PC
Motherboard
(Overclock)
Asus M4A77TD
Chipset: AMD 770/SB710
O/C to 254 MHz Ref. Clk.
MSI 790X-G45
Chipset: AMD 790X/SB710
O/C to 214 MHz Ref. Clk.
Gigabyte GA-X58A-UD3R
Chipset: Intel X58 Express
O/C to 195 MHz BCLK
Processor
(Overclock)
AMD Athlon II X3 435 2.9 GHz
Three Cores, 1.5MB L2 Cache
O/C to 3.556 GHz at 1.488V
Phenom II X3 720 2.8 GHz
Three Cores, 6MB L3 Cache
O/C at X4 to 3.4 GHz, 1.5V
Intel Core i7-930 2.80 GHz
Four Cores, 8MB L3 Cache
O/C to 4.1 GHz at 1.44V
Memory
(Overclock)
2GB Crucial DDR3-1333
CL 9-9-9-24, O/C at 1.60V
to DDR3-1355, CL 8-8-8-24
4GB Crucial DDR3-1333
CL 9-9-9-24, O/C at 1.50V
to DDR3-1426, CL 9-9-9-24
6GB Crucial DDR3-1333
CL 9-9-9-24, O/C at 1.60V
to DDR3-1560, CL 8-8-8-20
Graphics
(Overclock)
PowerColor Radeon HD 5770
800 MHz GPU,  GDDR5-4800
O/C to 910MHz, GDDR5-5640
2 x Gigabyte Radeon HD 5830
800 MHz GPU, GDDR5-4000
O/C to 820 MHz, GDDR5-4400
2 x Gigabyte GeForce GTX 470
607MHz GPU, GDDR5-3482
O/C to 643 MHz, GDDR5-3584
Hard DriveSamsung HD502HJ
500MB, SATA 3Gb/s
7,200 RPM, 16MB Cache
Western Digital WD3200AAJS
320GB, SATA 3Gb/s
7,200 RPM, 8MB Cache
Samsung HD103SJ
1TB, SATA 3Gb/s
7,200 RPM, 32MB Cache
OpticalSamsung SH-S223C
22x DVD±R, 48X CD-R
Lite-On iHAS124
24x DVD±R, 48X CD-R
LITE-ON iHES208-08
8x BD-ROM, 16X DVD±R
CaseCooler Master Elite 330Antec Three HundredAntec Three Hundred Illusion
PowerCM RS-500-PCAR-A3
500W, Dual 12V at 18A
Corsair CMPSU-650TX
650W, Single 12V at 52A
SilverStone DA750
750W Modular, 80-Plus Silver
CPU CoolerCooler Master Hyper TX3Cooler Master Hyper TX3Prolimatech Megahalems
Rev.B with Scythe Slip Stream
SY1225SL12LM-P fan
Software
OSMicrosoft Windows 7 Ultimate x64
GraphicsAMD Catalyst 10.5Nvidia Forceware 197.75
ChipsetIntel INF 9.1.1.1020


Overclocking requires extensive stability tests, which in turn consume huge quantities of time. Yet, there are legitimate shortcuts that can reduce the number of settings tested. For example, we’ve seen processors based Intel’s 45 nm production technology deteriorate fairly quickly (months rather than years) when pushed slightly beyond 1.45V, so the $2,000 machine’s CPU overclock stopped short of 1.45V, even though the machine had enough cooling to support the higher voltage levels needed to achieve even better clock rates.

Similarly, we’ve tested enough of these particular Crucial modules to know that they consistently exceed DDR3-1960 at CAS 9 and reach DDR3-1600 at CAS 8. With that tidbit of information, limited test time is the only somewhat-acceptable excuse today’s builders can use for not pushing the above modules farther into performance realms.

Benchmark Configuration
3D Games
Call of Duty: Modern Warfare 2Campaign, Act III, Second Sun (45 sec. FRAPS)
Test Set 1: Highest Settings, No AA
Test Set 2: Highest Settings, 4x AA
CrysisPatch 1.2.1, DirectX 10, 64-bit executable, benchmark tool
Test Set 1: High Quality, No AA
Test Set 2: Very High Quality, No AA
DiRT 2Run with -benchmark example_benchmark.xml
Test Set 1: High Quality Preset, No AA
Test Set 2: Ultra Quality Preset, 8x AA
S.T.A.L.K.E.R.: Call Of PripyatCall Of Pripyat Benchmark version
Test Set 1: High Preset, DX11 EFDL, No AA
Test Set 2: Ultra Preset, DX11 EFDL, 4x MSAA
Audio/Video Encoding
iTunesVersion:9.0.2.25 x64
Audio CD (Terminator II SE), 53 minutes
Default format AAC
Handbrake 0.9.4Version 0.9.4, convert first .vob file from The Last Samurai (1.0GB) to .mp4, High Profile
TMPEGEnc 4.0 XPressVersion: 4.7.3.292
Import File: Terminator 2 SE DVD (5 Minutes)
Resolution: 720x576 (PAL) 16:9
DivX Codec 6.9.1Encoding mode: Insane Quality
Enhanced multithreading enabled using SSE4
Quarter-pixel search
XviD 1.2.2Display encoding status = off
MainConcept Reference 1.6.1MPEG2 to MPEG2 (H.264), MainConcept H.264/AVC Codec, 28 sec HDTV 1920x1080 (MPEG2), Audio: MPEG2 (44.1 KHz, 2 Channel, 16-Bit, 224 Kb/s), Mode: PAL (25 FPS)
Productivity
Adobe Photoshop CS4Version: 11.0 x64, Filter 15.7MB TIF Image
Radial Blur, Shape Blur, Median, Polar Coordinates
Autodesk 3ds Max 2010Version: 11.0 x64, Rendering Dragon Image at 1920x1080 (HDTV)
Grisoft AVG Anti-Virus 9.0Version: 9.0.663, Virus base: 270.14.1/2407, Benchmark: Scan 334MB Folder of ZIP/RAR compressed files
WinRAR 3.90Version x64 3.90, Dictionary = 4,096 KB, Benchmark: THG-Workload (334MB)
7-ZipVersion 4.65: Format=Zip, Compression=Ultra, Method=Deflate, Dictionary Size=32KB, Word Size=128, Threads=8
Benchmark: THG-Workload (334MB)
Synthetic Benchmarks and Settings
3DMark VantageVersion: 1.0.1, GPU and CPU scores
PCMark VantageVersion: 1.0.1.0 x64, System, Productivity, Hard Disk Drive benchmarks
SiSoftware Sandra 2010Version 2010.1.16.11, CPU Test = CPU Arithmetic / MultiMedia, Memory Test = Bandwidth Benchmark
Display all 44 comments.
  • 6 Hide
    manitoublack , June 18, 2010 6:15 AM
    Another great SBM. Goodluck to US Punters who get the chance to win theses systems. Look forward to the next round where Graphics hardware will take a step out of the lime light.
  • 0 Hide
    wildeast , June 18, 2010 6:23 AM
    marry me tom's :D 
  • 3 Hide
    touchdowntexas13 , June 18, 2010 6:31 AM
    It's interesting to see the performance/dollar shoot up for the $2000 pc when it comes to games. That just goes to show you how much of the budget went into graphics muscle. These machines were definitely built with gaming in mind.
  • 0 Hide
    Crashman , June 18, 2010 7:14 AM
    touchdowntexas13It's interesting to see the performance/dollar shoot up for the $2000 pc when it comes to games. That just goes to show you how much of the budget went into graphics muscle. These machines were definitely built with gaming in mind.

    Yes, the only way to smash those benchmarks is with a faster CPU (2/3 of tests) or graphics (1/3 of tests). The problem with upgrading the CPU is that the 980X would cost 50% of the total budget. We haven't seen a big improvement in overclocking by using a higher-model quad-core i7
  • 3 Hide
    touchdowntexas13 , June 18, 2010 7:26 AM
    CrashmanYes, the only way to smash those benchmarks is with a faster CPU (2/3 of tests) or graphics (1/3 of tests). The problem with upgrading the CPU is that the 980X would cost 50% of the total budget. We haven't seen a big improvement in overclocking by using a higher-model quad-core i7


    Oh no I wasn't suggesting at all that you should have gone with a 980X for the $2000 build. That's way too expensive for a $2000 limit. The 930 does it's job just fine.

    It just amazed me that two 470's in SLI were able to best the performance/dollar of the cheaper builds. Typically you see diminishing gains as you get into the more expensive components.

    It was a very interesting set of articles any way you look at it. Gamers on a budget should especially be interested in this SBM.
  • 0 Hide
    Crashman , June 18, 2010 7:55 AM
    touchdowntexas13It just amazed me that two 470's in SLI were able to best the performance/dollar of the cheaper builds. Typically you see diminishing gains as you get into the more expensive components.
    I was pretty amazed too, but I really want to give credit to $1000 PC builder Don for making the GTX 470 SLI suggestion for the $2000 machine. Spot on Don!
  • 3 Hide
    Tamz_msc , June 18, 2010 8:46 AM
    Overall this month's SBM was good, especially the scalability of the 470s was brought into prominence.Though overclocking those in SLI is certainly not a viable option, unless one steals power from the Hoover Dam.
  • -6 Hide
    Willroo , June 18, 2010 9:20 AM
    Did anyone notice that the 858w microwave has a power supply rated for 750w.....sizzle.....pop.....anyone smell smoke...? Running f@h on that machine the power company would have to burn a ton of coal a day and you'd get threat mail from them when you cause a brown out. Ah....But all those PPD.
  • 4 Hide
    Onus , June 18, 2010 9:56 AM
    Interesting. Based on a previous article, an Athlon II X2 440 wouldn't be enough to let the 470s in SLI stretch out; I wonder what the minimum CPU there would be.
    AND, since many of us found problems in these builds, if those were "fixed" (possibly costing more), those results would be useful too.
    Lots of good information in this SBM round. Very nice.
  • 3 Hide
    Crashman , June 18, 2010 10:58 AM
    WillrooDid anyone notice that the 858w microwave has a power supply rated for 750w.....sizzle.....pop.....anyone smell smoke...? Running f@h on that machine the power company would have to burn a ton of coal a day and you'd get threat mail from them when you cause a brown out. Ah....But all those PPD.

    Silverstone says it outputs 77 to 80% of what you input. That's 670W of power output at 858W power input. It's rated at 750W continuous power output, but don't let the facts get in the way of a rant, eh?
  • 0 Hide
    dirtmountain , June 18, 2010 11:53 AM
    Nice work on all the builds with some interesting choices. I would never have thought that a 1,000w power supply would really be needed for those 470s. I enjoyed reading about every build, nice work by the SBM crew.
  • 1 Hide
    Onus , June 18, 2010 12:51 PM
    I'd like to win any of these. The first just needs a quality PSU, then I'd give it to a family member. If I won one of the others, I'd probably do some parts swapping with my current system(s), keeping the result as my primary system and giving the other one to a family member, and would likely have a GPU left over for another system.
  • 1 Hide
    Otus , June 18, 2010 1:16 PM
    Quote:
    Moreover, the shockingly power-hungry $2,000 overclocked system offers a similarly-surprising 95% efficiency rating compared to our $550 baseline.


    Image says 84%, which is rather less surprising.
  • -5 Hide
    Poisoner , June 18, 2010 2:06 PM
    Okay, we know Fermi uses a lot of power, big deal. So does the 4870. Damn Fermi jokes are already played out more than "Can it play Crysis?"
  • 2 Hide
    shovenose , June 18, 2010 2:13 PM
    i think the psu choices are a bit odd. the $550 needs a better psu, and the $2000 needs a bigger psu of the same quality...
    but they are great otherwise!
  • 1 Hide
    EnderYeah , June 18, 2010 4:03 PM
    I love these system builder marathon articles! Keep up the great work!
  • 0 Hide
    ta152h , June 18, 2010 4:11 PM
    I'm still not sure the 720BE is tested right. Unlocking a core could be completely misleading for someone buying the CPU, because they might not be so lucky, and then are going to be disappointed. It might be more meaningful if the core wasn't used. It's also a little scary running one since CPUs don't always fail dramatically, and sometimes just failed a small test that might not reveal itself for a while, and give bad results later on. I'm not that comfortable with that.

    But, perhaps more importantly, the 720BE might overclock better were it not for using the "bad" core. It's not that including the extra core results are so bad, but I think including the results without the core would have been very meaningful as well.

    It's almost contradictory to say you may or may not unlock a core, so beware, and then not include results in the event you can't unlock the core. Even with the uncertainty with CPUs in terms of overclockability, this adds more than is necessary.
  • 0 Hide
    rutoojinn , June 18, 2010 5:32 PM
    This SBM mainly the budget gaming system is great, shows you don't have to break the bank to get a pretty adequate gaming system. Of course it is hard to please everyone at every level but this one did a decent job. I would also like to see the 5770 in crossfire and how that performs and hopefully the Nvidia GTX460 will be out for testing also.
  • 2 Hide
    Pei-chen , June 18, 2010 5:48 PM
    I normally support Intel and Nvidia but I gotta say this test isn't fair. 5830 is the worst value in AMD/ATI's line up while 470 is Nvidia's best value.

    The $1000 would have a clear lead over the $2000 machine if it is paired with two 5770 or 5850.
  • -4 Hide
    misha87 , June 18, 2010 5:59 PM
    i have a pc that almost like about 700$ it is a quad core and it is even better than the 1000$ dollar one it is faster
Display more comments
React To This Article